8 Surprising Examples of AI in Security [NEWS]

From military reconnaissance to infectious disease prevention, here are 8 surprising tools and systems powered by artificial intelligence in security



Adam Muspratt
11/01/2018

The natural convergence of artificial intelligence and security 

Artificial intelligence can add value to virtually every area of society. The natural convergence of physical security, cyber security, and artificial intelligence has generated such a buzz that we are constantly seeing innovative and smart solutions emerge to face our most pressing security challenges.

Security is a broad concept and its meaning can change depending on who you are speaking to. From state security to private enterprise and even the individual, security will continue to be a chief concern as IoT connected devices become increasingly ubiquitous.

However, it’s not just the dissemination of private data that we have to worry about. Physical security still remains pivotal for the protection of confidential data, software equipment, facilities, and company assets.

Whether it is through advanced heuristics, machine learning or data analysis, artificial intelligence can offer solutions that are a less labor-intensive solution in comparison to the traditional method and can help smaller companies achieve a level of operational efficiency that is traditionally reserved for large corporations.

Cybersecurity is one of the most prominent cases uses for artificial intelligence and is an arena that is being explored by all actors. Artificial intelligence will likely serve as the foundation of cybersecurity procedure in the coming years, but what are other areas of security where artificial intelligence is becoming increasingly important? 

RELATED: An enterprise guide to ensuring IoT security

Here are 8 surprising things powered by artificial intelligence in security.

1. Underside vehicle bomb detection

UVeye, an Israeli startup has developed an artificial intelligence and machine learning product that helps security personnel detect threats by scanning the underside of passing vehicles.

Vehicle inspection system. Source: UVeye

The system uses strategically angled high-resolution cameras to create a 3D image of anomalies, such as improvised explosive devices, illegal weapons, drugs, and other suspicious material. UVeye claims that the technology will work even when the vehicle moving up to 28 MPH, meaning that the technology can detect objects that would otherwise escape the human eye.

RELATED: Transforming Defence Through AI and Machine Learning

This is where the artificial intelligence and machine learning comes in. The system has access to mountains of supplied data, ensuring that it can track the characteristics of the passing vehicle with manufacturer supplied data, and compare it with the data retrieved from the passing vehicle in real time. It can detect differentiators, such as part placement and even weight discrepancies. In addition, the system utilizes audio to listen for anything out of the ordinary.

According to UVeye, the technology takes up to three seconds to determine if a material is unlawful. Ultimately, this technology is applicable to a wide array of security roles and can change the way vehicle security is approached. This technology is applicable outside of government buildings, port facilities, border crossings, airports, power plants, tourist hot-spots and other critical infrastructure that might be an ideal target for an attack.

RELATED: What should governments be doing about the rise of artificial intelligence?

Outside of a national security setting, the technology can also be pivotal for the automotive market. The ability to detect commercial vehicle anomalies such as leaks and wear and tear damage can save consumers, garages and automobile manufacturers the time and money required to carry out traditional checks.

In the context of defense, this technology would likely see use at military checkpoints and areas where heavy traffic can be detrimental to safety. In fact, many defense experts believe urban combat will become more prevalent in the coming years, highlighting the need for the industry to develop proactive IED deterrence.

2. Infectious disease detection

From a public-health standpoint, infectious disease can be a huge problem – putting countless lives at risk as medical professionals battle to contain outbreaks. We only have to look towards the Democratic Republic of the Congo and the spread of Ebola to understand the severity of infectious disease outbreak. Ensuring that an outbreak prevention campaign can help the right people and the right time, and preventing the spread of disease is a major challenge for healthcare and humanitarian services.


Shutterstock

Fortunately, AI-based technology has been making great strides into the medical sector and companies are developing AI-based tools that can rapidly track, analyze and diagnose infectious diseases before it spreads

In September 2018, scientists from the National Environment Agency in Singapore created an algorithm to forecast dengue outbreaks, a mosquito-borne virus that affects over 400 million individuals.

In addition, Silicon Valley-based medical startup, AIME, has also developed an AI-based platform that is capable of predicting the exact geolocation data of infectious disease outbreaks like dengue or Zika up to three months in advance.

Check out our podcast with Ricardo Badillo from Western Union for an excellent discussion on RPA and information security 

The system predicts outbreaks by analyzing a combination of public health data, the weather, wind speed, proximity to bodies of water, and historical outbreak data. Essentially, it factors any variable that might affect the date, location, and severity of an infectious disease outbreak. This would be a particularly time-consuming effort for humans, in a situation which where time is not available.  

In addition to predictive countermeasures, AI-based technologies can a help with decision making through modeling. Traditionally, surveillance data of the outbreak is collected and public health officials consult with experts and stakeholders to design a containment and treatment programme. However, this isn’t always ideal as rapid decision making is difficult to achieve with so many stakeholders involved.

Ultimately, artificial intelligence is far more capable at pattern recognition and digesting mountains of data for the most optimal health outcome – and like other industries – data in healthcare is only becoming larger, with the growth of electronic health records, as well as digital unstructured data in the form of photos, genomic information, and health professional notes.

RELATED: Imagining a global security framework for artificial intelligence

At the moment, the exploitation of electronic health data via artificial intelligence is becoming a priority for fighting infectious disease. Artificial intelligence uses the most up to date data and research around the world, providing the fastest solution.

3. Home security

According to new market research, home security is expected to reach $ 74.75 billion by 2023, up from $45.58 billion in 2018. This rise is largely attributed to the introduction of artificial intelligence, deep learning, and the increasing ubiquity of IoT connected devices.

Indeed, artificial intelligence is revolutionizing home security as the technology can solve one of the biggest issues faced by traditional home security solutions; human error and false alarms. One solution, Lighthouse AI, uses artificial intelligence to alert you of humans, pets and other things that would be of interest while you're away from home. 

"Artificial intelligence is far more capable at pattern recognition and digesting mountains of data for the most optimal health outcome"

Using the Lighthouse app on a mobile device, the technology informs you of certain events, from a pet walking across your living room, or even a home intruder. The camera needs to be taught which people should be recognized as people who live in the house in order for it to function properly. 

Home security cameras have become incredibly powerful with the addition of artificial intelligence. Some traditional security systems would sound an alarm or alert the police if it detects movement. Obviously, this can result in false alarms if you or a family member enters the home without having primed the system to enable human presence at a particular time.

4. Threat screening for large events

Evolve Technology is an AI-based system that enables threat screening on a huge scale. The technology utilizes artificial intelligence and facial recognition software to analyze live footage of approaching visitors to determine if they are approved persons, such as regular visitors, VIPs, employees and other persons who should be granted entry.

Evolve Edge. Source: EvolveTechnology

If a visitor is highlighted as a non-permissible person of interest, their profile will be sent to security officers and a human expert can review and verify the data. The technology claims to allow at least one person to be allowed entry per second. This particular technology is not designed to completely eliminate the human element of threat analysis and can be best utilized at locations such as airports, sporting events and schools. If utilized successfully, this would effectively put an end to long lines and bottlenecks. 

5. Crime prevention cameras

Artificial intelligence and machine learning are adding a layer of proactive trouble detection to CCTV cameras. With the addition of artificial intelligence, CCTV cameras are now able to spot potential shoplifters and alert shopkeepers to suspicious behavior.


One solution called “AI Guardman” developed by a Japanese company can scan live video streams to form estimations of “suspicious” behavior. Through artificial intelligence and machine learning, the system tracks the posture and movement of shoppers, and analyses it to match the posture and movement of confirmed shoplifters derived from previous data.

RELATED: GUIDE TO AUTONOMOUS VEHICLES 

This technology is also being applied cameras outside of the home. San Francisco Company, Deep Sentinel, is developing a home camera system based upon a similar model. It works by utilizing algorithms to spot suspicious activity, such as an intruder planning to break in or neighbor trying to steal a package at the front door.

6. Military reconnaissance

The military potential for artificial intelligence is huge. There is a natural convergence between the two areas, with military hardware rife with cameras, sensors, communication networks, and data that would benefit from artificial intelligence. The capacity for humans to deal with the sheer volume of data on the modern battlefield is becoming a roadblock, affecting decision making and the ability for information to flow down to where it is needed the most.

Hivemind Nova. Source: Shield AI

US company, Shield AI has developed an AI-based surveillance solution for military personnel. The technology is driven by Hivemind, a machine-learning application that enables unmanned vehicles to progressively learn about the world around them. Effectively, this gives military units out in the field the ability to explore potentially life-threatening locations, such as building interiors, tunnels, and caves, with a substantially reduced risk to human personnel.

RELATED: AI 2020: The Future of Defence & Security

In practice, the drone can be commanded to survey a target environment, and the machine-learning technology onboard the drone will help it navigate without any human participation.

Ultimately, the inherent nature of defense means that sluggish data processing can result in increased danger for personnel. The Department of Defense has described artificial intelligence as the “most disruptive technology of our time” and is working on AI-based projects to improve decision-making cycles.

This is just the beginning. Decades into the future, as the technology matures, many defense experts have stated artificial intelligence will take a more active role on the battlefield both in the air and land domain. Future warfare may consist of ai-driven armored vehicle formations on land, and ai-driven pilotless aircraft in the sky.

7. Border control lie detector

Scientists from Manchester Metropolitan are working on a cutting-edge lie detector that can determine if a person is lying about who they are and why they are traveling.

iBorderCtrl – a European-funded project that has the potential to change the way we approach border control. Virtual border guard that uses artificial intelligence to tell if a person is lying or telling the truth through imperceptible signs, such as facial micro-gestures.



UK Border Control. Source: Shutterstock 

In practice, travelers are interviewed by an avatar border guard which asks a series of questions about their reason for travel, how they are traveling and more. The system takes the data and creates a psychological profile for the travelers and assigns a score based on their truthfulness. The system also utilizes machine learning, ensuring that it learns new methods of deception, ensuring that critical data to the project as each person uses the system.

"The Department of Defense has described artificial intelligence as the “most disruptive technology of our time”

Dr Keeley Crockett, a member of the European consortium developing the iBorderCtrl software at Manchester Metropolitan University, said: “It’s great to be part of such a huge and significant project, with partners across Europe, and it is exciting that research and technology developed in the heart of Manchester could revolutionize the future of travel and international security.”

RELATED: GLOBAL SMART CITIES IOT TECH REVENUES TO EXCEED $60BN BY 2026 

The technology is due to be tested at land borders all over Europe throughout 2018 – in Hungary, Greece, and Latvia. This demonstrates an increasing interest in smart city technology, as vendors from a wide array of sectors are repositioning their products to take advantage of the new opportunities presented by smart city technology.

8. Offshore Oil & gas threat detection

Artificial intelligence based solutions have also breached the oil and gas sector, a sector that has only fairly recently brought into the benefits of IoT connected devices and increased connectivity to the outside world. Indeed, artificial intelligence is revolutionizing how offshore oil & gas workers maintain the security of offshore oil and gas platforms.

Offshore Oil & Gas. Source: Shutterstock

There are currently several solutions on the market that utilize a combination of ai-driven software, sensors, real-time and historical data to assure cognitive security. For instance, offshore oil workers can receive recommendations and on how to prevent security breaches and ensure the longevity of equipment.

In the oil and gas sector, companies offer AI-based solutions for asset performance. With ai-based predictive asset solutions, oil & gas organization can benefit from early warning equipment and infrastructure failure, giving workers enough time to take measure and improve performance. This will help human workers cope with the enormous task of processing huge amounts of data and making safety-critical decisions in real time.

These systems work by assessing the operating profile of equipment during all working conditions – enough to create an advanced modeling process. The model is compared to the assets signature in real time, ensuring that deviations from critical infrastructure maintain responsible levels of stress.

This is pivotal in a sector that is comprised of critical and complex assets, with increasingly complex engineering systems which are becoming more interconnected and controlled by computers, including offshore drills, pipeline stations, and offshore oil & gas storage facilities.

If you enjoyed this article, you'll enjoy our podcast interview with Chris Gilmore from Aneta. Chris tells us how artificial intelligence is helping the medicare sector

OUR BENEFACTORS